Approximately half of children with acute myeloid leukemia (AML) eventually succumb to their disease. The long-term objective of the proposed research is to improve the clinical management of these patients (and hence their outcome) through sensitive identification of minimal residual disease (MRD) and tracking of leukemia stem cells. A flow cytometric method developed in the applicant's laboratory can distinguish 1 leukemic myeloid cell among 1000 or more normal hematopoietic cells and, in preliminary studies, was successfully applied in monitoring MRD in 85% of children with AML. A positive MRD finding after remission induction was the most powerful predictor of treatment failure.The studies proposed in this application are grounded in a multi-institutional study of AML, which specifies sequential MRD examinations in bone marrow and peripheral blood in approximately 200 children with newly diagnosed AML. Aim 1 is to extend sensitive MRD monitoring (i.e., 1 AML cell in 10,000 or more normal cells) to all children with AML by identifying novel leukemic cell markers. The proposed strategy relies on comparison of the gene expression profiles of AML cells (150 cases already studied) with those of normal immature myeloid counterparts to identify genes differentially expressed in AML. MRD detection by the newly identified markers will be tested against standard flow cytometric and molecular MRD assays. There is mounting evidence that AML is driven by a distinct subset of transformed hematopoietic stem cells, which are likely to be the most relevant targets for effective AML therapy. However, the clinical significance of leukemia stem cells in AML has not yet been systematically addressed. Thus, studies in Aim 2 will rely on newly-developed flow cytometric methods and on leukemia cell growth assays in vitro and in vivo to establish the prevalence of AML stem cells at diagnosis and determine the prognostic impact of their persistence. Aim 3 seeks to assess the clinical utility of MRD assay applied to peripheral blood instead of bone marrow. Based on preliminary results, the expectation is that improvements in the sensitivity of MRD assays will allow MRD to be identified in the peripheral blood of most if not all patients with bone marrow MRD, enhancing the prospects for more frequent MRD monitoring in children with AML. Studies in this aim will determine whether higher levels of blood MRD and of circulating leukemic stem cells correlate with a higher risk of relapse.